for long term investments, why not go 100% emerging markets?

[DcifrThs]

[ QUOTE ]
[ QUOTE ]

Montle carlo simulation are easy to understand and they are taught in intro business classes, or if you used pokerstove. They apply perfectly to this situation.

[/ QUOTE ]

sure it applies, but for the point he was making theres easier ways to get the point across. using the monte carlo option on pokerstove doesn't equal understanding monte carlo simulation. but whatever, thats not really the point of this thread.

[/ QUOTE ]

if you read my post, you'll see that after the paragraph where i mention "monte carlo" (much to your dislike) i state unequivocally that the above paragraph ISN'T the most important part and is just a thought exercise to show what could happen as you approach the end of your time horizon. the most important stuff is what proceeds the "monte carlo" discussion.

wrong side of the bed this morning?

Barron

[stinkypete]

[ QUOTE ]

if you have a portfolio that returns 15% at 45% volatility with no leverage, you are equally likely to be "wiped out" if the exact same investor has a 5% return portfolio with 15% volatility leveraged 3:1. (i picked 5% return here to make the resulting portfolio identical save transaction/leverage costs)

[/ QUOTE ]

that's only true if you stay leveraged exactly 3 to 1, which would mean continuously buying or selling whatever youre invested in. but adjusting your positions continuously is not possible, and even if you do adjust on a regular basis, your costs will be much higher.

if you're leveraged 3 to 1 and your portfolio drops 20%, you now only "own" 13.33% of your portfolio and youre suddenly leveraged 6 to 1. if it drops 33% before you make adjustments, you've gone broke. you'd likely make adjustments before that happens, but this illustrates why it's not the same. if you're leveraged 1 to 1 and the same 33% drop occurs, you've only lost 33%.

modeling returns with an expected value (average return) and a standard deviation (volatility) really doesn't work with lots of leverage. it's too far from a normal distribution.

and yes, i was hung over this morning.

[DcifrThs]

[ QUOTE ]
[ QUOTE ]
another way to look at it is to imagine a monte carlo simulation where you run the "lumpy" investment 100 times for 20 years vs. the "flat" investment 100 times. ni the lumpy scenario, there are instances (maybe 5% or something) where the last year is a -10% or -20% year that wipes out a large amount of the previous 20 years of gains. the flat scenario would almost never have that huge loss, but wouldn't have those huge gains also.

[/ QUOTE ]

it baffles me every time i hear people who are actually educated in these matters say something like this.

who cares if you drop 20% in the last year? you could also drop 20% in the first year, wiping out 20% of your gains for the next 19 years.

guess what? multiplication is commutative!

consider this example:

what is 0.8 * 1.15^19?
how about 1.15^19 * 0.8?

and if you suddenly become risk averse in your last year, you can always invest in something else.

also, i don't see why you're throwing around jargon like "monte carlo simulation", which half the readers here won't understand, when you could easily get your point across using simple explanations.

[/ QUOTE ]

the point is as was described in the above posts: that last .8 hurts more than the first since you do become more risk averse as time goes on (or so studies on that matter show). a simple example: think about how you plan for your retirement in the .8*1.15^19 scenario. is your retirement plan suddenly in jeopardy? not really since you can expect a 20% down year and you knew that when you went for the 15% return.

but what about the 1.15^19*.8...now it does have a very large impact since you've planned for presumably some # of years greater than 1 on that retirement income and BOOM all of a sudden you have a big loss in the last year that you didn't plan on.

so sure, mathematically they are the same. but in terms of investing and living, they are not.

anyways, the whole point of this thread is imo to consider risk when investing in things that look attractive from an absolute return standpoint.

i even said in the very next paragraph that it was a thought exercise and doesn't prove anything useful. the bulk of my post dealt with the important stuff in this thread.

Barron

[stinkypete]

[ QUOTE ]

but what about the 1.15^19*.8...now it does have a very large impact since you've planned for presumably some # of years greater than 1 on that retirement income and BOOM all of a sudden you have a big loss in the last year that you didn't plan on.

[/ QUOTE ]

well, if you're in highly volatile investments you shouldnt be making plans like that. if you're going to make those plans, you're going to reconsider your positions if you're not a dumbass.

but that's no argument for investing like a nit for the first 15 or so years. yes, you should be trying to maximize your risk adjusted returns - within reason. over a long time period, you should first make sure you're maximizing your returns and worry about lowering your volatility second - within reason again, of course. leverage isn't always that easy to come by, particularly in retirement accounts. it's not as simple as finding a portfolio that maximizes your sharpe ratio and then leveraging the crap out of it. over a long time period, usually the portfolio that makes the most sense won't maximize your risk adjusted return (though it should come close), because it's impossible to get the returns you want with that particular portfolio because you can't leverage it easily.

[blendedsuit]

[ QUOTE ]

because it's impossible to get the returns you want with that particular portfolio because you can't leverage it easily.

[/ QUOTE ]

leverage 3:1? where are you going to do that? besides forex

[DcifrThs]

[ QUOTE ]
[ QUOTE ]

but what about the 1.15^19*.8...now it does have a very large impact since you've planned for presumably some # of years greater than 1 on that retirement income and BOOM all of a sudden you have a big loss in the last year that you didn't plan on.

[/ QUOTE ]

well, if you're in highly volatile investments you shouldnt be making plans like that. if you're going to make those plans, you're going to reconsider your positions if you're not a dumbass.

[/ QUOTE ]

i'm using extreme examples here to get the point accross. the simple point: timing of returns makes a difference in life whereas it doesn't in math.

[ QUOTE ]


but that's no argument for investing like a nit for the first 15 or so years. yes, you should be trying to maximize your risk adjusted returns - within reason. over a long time period, you should first make sure you're maximizing your returns and worry about lowering your volatility second - within reason again, of course. leverage isn't always that easy to come by, particularly in retirement accounts. it's not as simple as finding a portfolio that maximizes your sharpe ratio and then leveraging the crap out of it. over a long time period, usually the portfolio that makes the most sense won't maximize your risk adjusted return (though it should come close), because it's impossible to get the returns you want with that particular portfolio because you can't leverage it easily.

[/ QUOTE ]

in reality, you don't choose between a portfolio with 15% returns and 45% vol vs. 12% returns and 15% vol.

instead, you often have choices like "should i invest all my money in EMD & EMEquties since they'll outperform over 20 years?" or "should i keep all my funds in a diversified stock portfolio with some bonds in it?"

the point i'm trying to make is that you can do better than both of those with a SMALL amount of leverage. 3:1 on 12%ret w/ 15% vol was an extreme example to show a point.

and over the long run, given you have access to a small amount of leverage, the best passive portfolio is the one that maximizes your sharpe ratio. that portfolio will dominate all others.

if you don't have acccess to any leverage, then sure, you can have an equity heavy (or emerging market heavy) portfolio since that will give larger absolute returns over time than a totally diversified portfolio. but as time passes (and your retirement nears), you should look to add lower vol lower returning allocations. this is very easy if you have a yearly contribution thing going on since you can just allocate a little bit to those assets as the time comes.

Barron

[Groty]

People have a tendency to extrapolate current conditions into the future. We saw it in 2002 when risk aversion was so extreme that bids vanished causing credit default spreads to widen ot near record levels. This past spring, only five short years later, we observed the exact opposite psychology where risk seeking behavior was insatibale resulting in the tightest credit spreads ever for certain rating of non-investment grade credit.

From John Mauldin:

"George Friedman of Stratfor points out (in a manuscript I just read for his new book, due out soon, we hope) that as humans we are dismal at projecting the future in a geo-political sense. Think about the beginning of every decade of the last century. Then move forward 20 years. Who got any prediction right? Did anyone see World War 1 in 1900 or even 1910? The rise of Germany and Hitler in 1920? The collapse of Russia in 1980? The rise of radical Islam in 1987? A war in Iraq in 2000?

And yet, we assume that the world of 2017 or 2027 will be not all that much different than today. But the only truly safe bet is that it will be radically different in ways that we cannot imagine."

[gull]

The reason that early losses are tolerable and late losses are less tolerable is that you do/can work and contribute capital to the portfolio. If you're 25 with a $25,000 portfolio, a -20% return is tolerable because you'll still be working and contributing. If you had a hands-off static portfolio, then losses at the beginning and end would be equivalent, as many of you have mentioned.

[PartyGirlUK]

What are some good long term investments other than emerging markets. Id like to have something high return/risk as well as E.M, but obv dont want all eggs in 1 basket.

[DcifrThs]

[ QUOTE ]
[ QUOTE ]

if you have a portfolio that returns 15% at 45% volatility with no leverage, you are equally likely to be "wiped out" if the exact same investor has a 5% return portfolio with 15% volatility leveraged 3:1. (i picked 5% return here to make the resulting portfolio identical save transaction/leverage costs)

[/ QUOTE ]

that's only true if you stay leveraged exactly 3 to 1, which would mean continuously buying or selling whatever youre invested in. but adjusting your positions continuously is not possible, and even if you do adjust on a regular basis, your costs will be much higher.

if you're leveraged 3 to 1 and your portfolio drops 20%, you now only "own" 13.33% of your portfolio and youre suddenly leveraged 6 to 1. if it drops 33% before you make adjustments, you've gone broke. you'd likely make adjustments before that happens, but this illustrates why it's not the same. if you're leveraged 1 to 1 and the same 33% drop occurs, you've only lost 33%.

[/ QUOTE ]

not quite. imagine two portfolios A & B

A: $1000 exposure on $333 leveraged 3:1
B: $1000 exposure on $1000 leveraged 1:1 (i.e. no leverage)

both lose 20%. the result?

A: $800 exposure now on 266.4 leveraged 3:1
B: $800 exposure now on $800 leveraged 3:1

so as you see, the only thing that changes is your margin. the portfolio naturally adjusts. losses on the portfolio are proportional to the equity upon which it is leveraged.

[ QUOTE ]


modeling returns with an expected value (average return) and a standard deviation (volatility) really doesn't work with lots of leverage. it's too far from a normal distribution.

[/ QUOTE ]

well that is an entirely different issue. there are numerous problems with the normal distribution when used on non-normally distributed data and thus large issues using it to model financial return streams.

the problem overall is clear, but the solution less so. when talking about return targets and volatility, it is still instructive to talk in mean or expected returns and volatility.

[ QUOTE ]
and yes, i was hung over this morning.

[/ QUOTE ]

lol, apology accepted [img]/images/graemlins/tongue.gif[/img]

Barron